pouët.net

This might be obvious, but I think it's important, so I'll make it explicit: Please try to spec out the API thoroughly and get Gargaj involved in the speccing before writing any code for it. Once the API goes live in v0.9 3rd party sites will start depending on it, and as soon as we switch to v2 (and I, too, trust Gargaj to deliver on his promise) Gargaj's code will have to support whatever warts were specced into the API.

My two cent.

"dipswitch: sure, tell analogue to stop the 1.0, the gargaj 2.0 will be way better !", yes it's the opposite, having one man on one code base has always been wrong. sometimes there is no choice, but... not here with a community, dip may help as well. Why wait years to say "wow 2.0 is bullshit, I'm going to make 3.0, but you can't look at it till I release it and then I'll fix it with a beautiful thread"... no point.
API is cool.

dojoe: that's good feedback and matching the feeling I have.

What about:
1. OSS v1.0
2. Spec a REST API
3. Build the REST API in v1.0

While this happens:
1. Gargaj finish up v2.0
2. Build the REST API in v2.0 (should not be that diff with the 1.0 code)
3. OSS v2.0

And finally, we switch pouet to v2.0

How's that ?

i think all this should be communicated by email between you and gargaj.

As Garg said he's not against opensourcing, but in my opinion the "I'm back" attitude and unwillingness to listen and cooperate understandably is rocking the boat. How would you feel if a project you took over for years was suddenly pulled out from under you by someone who hasn't been around for years?
Gargaj is one of the most reliable and considerate people I know, especially when it comes to the Demoscene. He will NOT go back on his given word, and he's not someone who'd just go and claim ownership of something like pouet 2.0 - due credit of course, but that's a given. From what I've seen up close he's been doing this as a service to the community and never expected anything in return, and this won't change.
Sit down and talk things through properly please before this turns into a bigger scene drama than needs to be. I know he's willing to listen if the attitude is right.

100% agreed.

Well we have an email thread, but he wanted things to happen here. I'll switch back.

mog, dipswitch, boyc: i don't see how inciting an ownership war over pouet will help anyone. analogue still has access to 0.9.x and full power to alter it, he wants to open source it, that's his decision. i understand your concerns thought.

i must remark i have not been in talk with analogue at all since my previous similar topic on this issue. i was happy to see this topic pop up though.

like you all i don't want to see 2.0 delayed.

answering some specific concearns:

[quote]Why not finish Pouet 2.0 first and then open sourcing Pouet? "
we would be depending on gargaj+tomoya to finish it soon (tm) (c) and then for them to be willing to open source their work. gargaj seems resistant to the later. so it could quite possibly all just end up in more months of waiting around and then the community at large not being able to contribute. by open sourcing what exists people have the option to contribute now.





i think everyone agress with the first part, that the codebase is crap. but there are a number of people willing to have a go at upgrading it. so why not enable them to do so while we wait for 2.0? analogue says he will control the merges and pull requests, so a vision is still in place.



for starters most of the patches will probably be small cosmetic fixes and idea implementations. everyone knows 2.0 and demozoo is in the works. so i seriously doubt anyone would want to venture into a complete db overhaul without checking with gargaj if it makes sense to just wait or how it could be also easily incorporated into 2.0 later.

as far as i know the most problematic difference between versions is the db. considering it will be barely touched and all the changes will be documented publically, i fail to see how this will delay any work on 2.0.
in fact, anyone wanting to contribute to pouet 2.0 or demozoo will now be able to tinker with pouet code to facilitate db access / migration.
the way i see it it's actually an advantage.

if gargaj would share his db migration plans people could even help him do it.

and then there is work being done on the API, which is a new approach to access the data, and i don't see why that can't be emulated / repurposed for 2.0 or any other site when it's live.


the most obvious positive point is that whenever you see something on pouet that has been annoying you for years, and you can't wait for demozoo or pouet 2.0, you can now just submit a patch to fix it.

I think whoever said this really missed what was actually said. As far as I can tell, Gargaj is perfectly willing to let this go open source when it is ready. With modern project hosting sites such as Github, Bitbucket, etc. it is incredibly easy to open source this: just turn off the private flag on the repository.

Personally, though, I really don't give two expletives whether or not the next version is open source. I write code, but that doesn't mean I have interest in adding any to Pouet (no offense, really). There are going to be a lot of people like that and a chunk of people clamoring for it in this thread probably wouldn't bother either. So what does open sourcing 2.x buy the authors? On top of that, who would actually bother to set up their own site based on the Pouet code? Probably not enough to make that a main driver either.

I think it's great that someone is actually doing something. Time taken is irrelevant. Most of us have work, and other things to keep us busy. It's hard to continuously work on something of this magnitude consistently.

(I have entered the mud pit!)

I should also note that I've been on the receiving end of notices about "bugs" and so on for things not yet completed or push to finish "new" things when other stuff hasn't been finished yet. Sometimes it's really just unintentional on the part of the person telling you, but can definitely wear a little thin. Occasionally I've gotten well-meaning notices about something in code from someone who really has no idea about that language or what is going on.

So for the developer's sanity, I think there's quite a good reason then to try and stave this off by not putting anything out there until it's stable enough. Then what gets reported is genuine, what you really want to know.

(A side note on that is I don't think Github, Gitorious, Bitbucket, etc. allow private projects with public ticket trackers. So the other benefit of waiting is the ability for people to easily enter new tickets than having the developers waste their time proxying them from multiple sources back into a private tracker.)

This thread made my day: http://www.pouet.net/topic.php?which=8683&page=1

I'm really happy to not to share a workplace with some of you.

OpenSourcing .9 will lead to new "features", thus people expect them to be in 2.0 as well. Good luck in keeping two codebases API similar, works like a charm for Perl for years, not.

What's so hard in asking Gargaj where you can help with the new one? And sure you have your bragging rights in starting this mess, but Gargaj and Tomoya seem to have more clue on what needs to be done than you. There's no need in rushing something out the door, that partially isn't even your codebase anymore and hope the bugfairy will fix it for you.

Good job, you just explained why you shouldn't be head of the project.

Oww gawd, you wanted the discussion and now people disagree with your "idea", we surely are all haters not knowing what we speak at all.

YES! Let's code the same stuff twice, because nobody on your SW team would facepalm over that - right? RIGHT? I also can see how we all can benefit of a REST API where the encoding is b0rked and we can shuffle more funny encoded things into the DB.

Analogue: If you really want Gargaj to opensource 2.0, I seriously doubt undermining all his work on it up until now is the right path. And yes, that is actually what you seem to be doing.

I get that you don't want the code to be "owned" by anyone and that you want an easy way to contribute. But if you had read the other thread you liked so much very closely you would have seen that Gargaj has said that he would like to finish the basic structure of 2.0 before considering opening it up.

And why? Could it be that he spent a lot of time wrapping his head around the system and thinking it out? Working on it on da day-to-day basis? He probably knows it by heart. Way more than anyone else here who happens to have access to the code. That includes you, you said it yourself when you said you were tired of spending ages trying to figure out what was going on before you could do anything. Why not let him finish his vision and plans solidly and THEN make a case about opening it up?

I also get that you're tired of waiting for it, a lot of people probably are. But again, why the rush at this late stage? Why now? Why not trust the person who has run your project for years while you've been away doing "fun things"?

Is it so damn hard to stop and think that maybe these people who make that advice might be right even if it's not what you wanted to hear when you first tried to ask what people thought?

Is it so damn hard to contribute and discuss and listen to others instead of just jumping in and managing things top down because you suddenly have some spare time?

And yes, I know it's your brainchild, you deserve some respect for that. But Gargaj deserves respect too. To use your own phrase: deal with it.

Please make sure 2.0 allows for flamewars and massive amounts of time being wasted on debating hypothetical issues too ;-)

mog,leia: how can we help Gargaj with 2.0 when it's closed and private? i'm sure he has the best intentions in mind, but hogging a project development of public interest for years on end is no good for anyone.

it's great that his version has a solid project leader, but there is no way for anyone to contribute except to wait. and waiting is what we been doing for years and years.

props to him for the work he's supposedly been doing but i don't see why the rest of us should all sit and wait and hope for the best or "go code a demo instead while i'm too busy" just because he thinks so.

you all talk about respect, but i'm seeing none of it. just fuel for ownership quarrels. it's just a matter of 2 guys, analogue and gargaj talking to sync roadmaps. i fail to see what ownership and respect has to do with it. both of them want 2.0 to come as soon as possible and with the most support as possible. so just let them talk and figure it out. and i just hope they won't go into ego ownership bullshit and ruin pouet even more otherwise both of them will get quite a earful out of me.

ps: which part about "gargaj wants to finish this part of 2.0 on his own because that's the best way to ensure that the foundation is solid" did you miss?

Note again the "solid foundation" part. If I read people correctly, that's been the biggest problem of 0.9/1.0 (especially the database, it seems). Throwing more people into that right now just seems like a good way to create a mess that is more of an hinderance than helping the project forward. Especially when the coder said so. If you want to help Gargaj, why not listen to what he says and respect his wish?

Like plenty of people said. Solid foundation, then discuss opening it up and contributing.. Why are you so against that?

Also, I undestand the impatience, but after all, you did get a link to the beta version yesterday (if not before like plenty of people have) so you can see for yourself that work has been done. So stop the "supposedly been doing" bullcrap, it's below you.

I simply asked him.

I wrote a couple of greasemonky scripts, what did you do?

So let's use your API http://codepen.io/mog/pen/KtnvF, well - bummer - doesn't work - who'd have thunk?

Additionally please compare pouet-api with what Pouet already does..

This discussion reminds me of every single open source project I've been somewhat involved in. I have since become an advocate of proprietary software.

Analogue:
Ok. So you are more available lately but the "oldschool" workflow of Pouet 0.9 is a hurting any progress. I feel your pain. I've been in these shoes when I tried to work on Pouet 0.9 a few years back.

I see three reasons why the respective admins of Pouet have kept things this way:

1. it did the job... somehow.
2. no access to a DCVS with private repositories
3. to focus on Pouet 2


psenough:To me it was clear that Pouet 2 includes a DB "refactoring". Also Gagaj said earlier that the current, important, tasks in progress on Pouet 2 are too specific to derail and put more people on.

As mog said, the community have written and published a bunch of user.js and css to bring such little improvements for quite some years now. AttentionWhoreBlocker, YTEmbed, ScreenshotPreview, UnicodeOnTheOnliner, LivePostPreview, ...

Gargaj, how about this: Put the 2.0 source on Github/whatever and state loudly and clearly (as in first line of readme.md) that you won't accept any pull requests until it's reached live status.

It has the disadvantage that suddenly people will have opinions and all that annoying stuff (and I'm the first one to admit that it isn't any fun - every time I opened up V2 a bit more the questions/demands got stupider).

But, seriously: If there's one thing inside the demoscene that really should be a community driven effort it's Pouet. As many flaws this place has, people care about it, and already there's more contributors than with your average OSS project. And we've got quite a selection of development professionals in here. As I see it, there's two advantages:

- If people outside can actually track progress (instead of just knowing "yeah, I heard about a pouet 2.0 in 2008") it will breed confidence even outside the crowd that already knows that you're reliable.
- As annoying comments on an unfinished versions are, there's a chance that somebody might raise a very valid red flag (eg. in the db schema or your object model) and thus prevent the suck from creeping into the new foundation in time. Early code reviews are a GOOD thing, especially when we'll have to live with that code for the next 12 years. :)

leia: none of it.

mog: only the whole pouet code from the point since analogue abandoned it, up until i ran out of time and patience to deal with it and passed it to gargaj. i know it needs a proper rehaul. keeping it's plans in a vault ain't accomplishing it though.

p01: and how many people are actually using those scripts i wonder.

also, what kb just said.

there is a middle term somewhere, it's only a matter of analogue and gargaj _talking_ about it and reaching a compromise. as opposed to swapping some emails on how each wants to go their own way and everyone else fueling an ownership feud on the forums.

I still think that the discussion is a bit pointless with demozoo in the picture.

Having two separate, unsynced release databases is a far bigger problem than "pouet needs to have a new icon and it needs it NOW".

or we need to come up with a good name for a fork! :P

How about PooIt for the .9 branch analog is desperately trying to make happen.

pouet using the demozoo prod db and adding the social drama on top of it would of course be a wise decision.
but i guess too many work has been put into stuff that would be almost obsolete then to really think about that move.